> "TikTok empowers users with transparent information about its privacy practices and gives them multiple tools to customise their experience," a TikTok spokesperson says. "Advertising pixels are industry standard and used widely across social and media platforms"
Such Doublespeak—the word empower really means enfeeble and privacy its opposite.
Facebook was doing tracking pixels in the 00s. It probably worked even better then because stuff that's currently in apps was on the web back then and fewer people ran adblockers.
Every party in the advertising ecosystem should be assumed to be doing this (and your adblocker should be trying its best to block it).
"TikTok" in the headline for views but every ad system is sucking up as much data as it possibly can: cross-site tracking pixels, cookies, device ids, fingerprinting, app snooping, extension snooping, etc.
I hadn't read that Paul Graham article before, but it was extremely accurate at the time.
My degree is in Public Relations and I worked in political PR for a bit before moving to newspapers. The PR office worked so hard to word things in a way where news editors could lift our copy directly into print. It was a delicate balance to sell a point of view without sounding like a sales pitch.
Later, at the newspapers, I was shocked to learn how desperately editors would snag any text to fill the space between paid-for ads on a page. A minimal amount of actual journalism occurred above the fold. Past that we would publish absolutely anything in the English language without filtering.
This was all 20+ years ago. Now we've cut out the middle man, automatically publishing AI generated slop directly as if it were human-produced news. It's all very discouraging.
Paul Graham cherry picks examples of history where humans are comically misguided as an argument for why virtually all humans are copying each others morals. Graham somehow argues that stopping the Nazis in WW2 or protecting children by socializing them against risky activities is unthinking hive mind and hand waives all of the nuance and human integrity. Weak argument and reads like propaganda from Steve Bannon.
TikTok is now a Zionist operation being run by (former?) members of Mossad's Unit 8200, which is like the NSA's cybersecurity group. So monitoring everyone is literally the point of TikTok now. Meta, Google, Apple, and others are also participating in it. Silicon Valley not actively mobilizing against this shows how geeks are complicit with genocide and the systems that drive it.
TikTok, more than any other app, seems to be aware of things that I talk about. I'm not big on conspiracy theories (well until the past six months or so), but I really wonder if TikTok has figured out a way to listen with the microphone on my phone. I will be chatting about the most random thing -- needing a new washing machine -- and then I'll suddenly get some washing machine add in the next hour. Or someone will mention a movie being snubbed for the Oscar's, and then an edit for that movie pops up.
I never did a search or anything else on any app on any devices related to these things, but somehow TikTok seemed to know. Maybe coincidence that I have heightened awareness of... but it does seem different.
This wasn't a problem until it was done by a Chinese company, when American companies (Meta, X, Google, etc.) spied on us we saw it as a triumph of entrepreneurism.
Most popular platforms are tracking and spying on you. My friends and I also believe Slack private DMs are compromised as we often times see ads directly pertaining to oddball discussions we don't have outside of Slack.
Most people here probably know this already, but you can minimize some of this by using privacy browser extensions [1], containerized browsing [2], a good VPN [3], and/or Pihole [4].
The headline makes it sound uniquely sinister, but most of what's described here is just the modern adtech stack doing what it's been doing for a decade. The real tension is that advertisers want attribution, sites want revenue, and users want privacy and the current system optimizes almost entirely for the first two
Nothing new here. This is why they eventually rolled back Chrome's initiative to automatically reject third-party cookies. Industry backlash was that the analytics of too many sites would break. Best thing to do is to switch to a privacy centric browser.
I even read that TikTok has its app listen to a port on localhost, and have websites run code that exfiltrates data this way (effectively bypassing privacy protections of your browser).
(b) display images. If images are optional, then the option must be enabled, e.g., by default
For example,
I often fail these requirements as I manually retrieve information without a browser. For example, I read the information from the BBC website without meeting the software requirements for social media pixel tracking
I also use a text-only browser to read HTML offline. This browser fails the software requirements as it does not auto-load resources. Further, I compile it without support for images
In addition to the software requirements there is also a requirement for access to remote DNS controlled by a third party
If you do not use the TikTok website, then your browser has no need to retrieve DNS data for tiktok.com or other domains registered to or used by TikTok
Unless you delegate lookups to a third party DNS provider such as an ISP, Cloudflare, Google, Quad9, etc. or run a local resolver that accesses remote authoritative servers then the required web browser specified above will not be able to retrieve DNS data for tiktok.com or whatever domains are used for the tracking pixel
For example, I use only locally-stored DNS data served from local authoritative DNS servers and localhost forward proxy memory. There is no DNS data for tiktok.com or other domains used for TikTok's tracking pixel
NB. The subject of this comment is (c) software and DNS requirements for pixel tracking. A different subject is (d) how many users may or may not meet such requirements, e.g., high numbers versus low numbers, "average" users versus non-"average" users, and so on. HN replies often attempt to change the subject to (d)
40 comments
[ 2.8 ms ] story [ 52.8 ms ] threadSuch Doublespeak—the word empower really means enfeeble and privacy its opposite.
https://github.com/danhorton7/pihole-block-tiktok/blob/main/...
Every party in the advertising ecosystem should be assumed to be doing this (and your adblocker should be trying its best to block it).
> For more science, technology, environment and health stories from the BBC, follow us on Facebook and Instagram.
If you worry about tracking you certainly shouldn’t be on Facebook or Instagram, one the OGs of pervasive tracking even if you’re not a user.
https://www.theguardian.com/technology/2022/mar/30/facebook-...
Also relevant: https://paulgraham.com/say.html
My degree is in Public Relations and I worked in political PR for a bit before moving to newspapers. The PR office worked so hard to word things in a way where news editors could lift our copy directly into print. It was a delicate balance to sell a point of view without sounding like a sales pitch.
Later, at the newspapers, I was shocked to learn how desperately editors would snag any text to fill the space between paid-for ads on a page. A minimal amount of actual journalism occurred above the fold. Past that we would publish absolutely anything in the English language without filtering.
This was all 20+ years ago. Now we've cut out the middle man, automatically publishing AI generated slop directly as if it were human-produced news. It's all very discouraging.
Bold claims like this need citations.
But this article is from the BBC so they would arguably be talking about the one that's available in Europe which is still controlled by China.
I never did a search or anything else on any app on any devices related to these things, but somehow TikTok seemed to know. Maybe coincidence that I have heightened awareness of... but it does seem different.
Most people here probably know this already, but you can minimize some of this by using privacy browser extensions [1], containerized browsing [2], a good VPN [3], and/or Pihole [4].
1: https://duckduckgo.com/compare-privacy?tab=extensions
2: https://support.mozilla.org/en-US/kb/how-use-firefox-contain...
3: https://docs.google.com/spreadsheets/d/1CRtEQzSVE59jj5ROKZlt... (*do your research, e.g. NordVPN creeps me out with aggressive advertising practices even though they're highly rated)
4: https://pi-hole.net/
Oh wait no sorry, that was Meta. In both the Facebook and Instagram apps: https://cybersecuritynews.com/track-android-users-covertly/.
You must use a web browser
The web browser must
(a) auto-load resources, e.g., images
(b) display images. If images are optional, then the option must be enabled, e.g., by default
For example,
I often fail these requirements as I manually retrieve information without a browser. For example, I read the information from the BBC website without meeting the software requirements for social media pixel tracking
I also use a text-only browser to read HTML offline. This browser fails the software requirements as it does not auto-load resources. Further, I compile it without support for images
In addition to the software requirements there is also a requirement for access to remote DNS controlled by a third party
If you do not use the TikTok website, then your browser has no need to retrieve DNS data for tiktok.com or other domains registered to or used by TikTok
Unless you delegate lookups to a third party DNS provider such as an ISP, Cloudflare, Google, Quad9, etc. or run a local resolver that accesses remote authoritative servers then the required web browser specified above will not be able to retrieve DNS data for tiktok.com or whatever domains are used for the tracking pixel
For example, I use only locally-stored DNS data served from local authoritative DNS servers and localhost forward proxy memory. There is no DNS data for tiktok.com or other domains used for TikTok's tracking pixel
NB. The subject of this comment is (c) software and DNS requirements for pixel tracking. A different subject is (d) how many users may or may not meet such requirements, e.g., high numbers versus low numbers, "average" users versus non-"average" users, and so on. HN replies often attempt to change the subject to (d)